Fast Power system security analysis with Guided Dropout

Source

• Raw Markdown: guided-dropout-power-system-2018
• Rendered / retrieved PDF: paper_guided-dropout-power-system-2018.pdf
• External source: https://arxiv.org/abs/1801.09870
• Official L2RPN reference list: https://l2rpn.chalearn.org/papers-references

Publication And Credibility

• Paper date: 2018-01-30.
• Venue/status: ESANN 2018; arXiv preprint available.
• Credibility: Historical RTE/ChaLearn power-grid ML source. Older than one year; use as lineage for fast security-analysis surrogates.

Core Claim

The paper uses guided dropout to predict power flows across topology variants while training on limited configurations.

L2RPN / Grid2Op Notes

It targets fast security analysis, where exhaustive physical simulation of every topology or contingency is too expensive for real-time operation.

Action-Time-Series Notes

This source is useful when Grid2Op is treated as an action-conditioned graph time-series environment:

power-grid observations + topology / redispatch / storage control input + scenario context
  -> next grid observations + safety/cost outcome

The terminology distinction matters. Topology changes, redispatching, curtailment, and storage commands are actions or control inputs when an agent chooses them. Line failures, maintenance outages, weather-driven renewable shifts, and demand variation are events or exogenous variables unless they are deliberately controlled by the experimenter.

Foundation TSFM Relevance

Agenda slot	Verdict	Evidence	Missing pieces
Causal structure, counterfactuals, and control	partially closes	Relevant to TSFM benchmarks because learned surrogates must generalize across structural/topology changes, not just interpolate time windows.	It is not a sequential agent benchmark and has no explicit reward/action rollout loop.
Context interface: topology and channel context	partially closes	Power-grid state is naturally graph-structured and tied to physical assets, limits, and scenario metadata.	Needs a reusable schema that a general TSFM can consume across grids and non-grid operational systems.
Benchmark level	adjacent	L2RPN/Grid2Op provides simulator-backed trajectories with explicit controls and outcomes.	TSFM-ready comparisons require pinned environment versions, action sets, reward definitions, and train/test scenario splits.

Links Into The Wiki

• Grid2Op
• Action-Conditioned Time-Series Datasets
• World Models
• Foundation Time-Series Model Research Agenda
• Introducing machine learning for power system operation support
• Anticipating contingengies in power grids using fast neural net screening
• Optimization of computational budget for power system risk assessment
• LEAP nets for power grid perturbations
• Graph Neural Solver for Power Systems
• Neural Networks for Power Flow: Graph Neural Solver